Interpretability Researcher. Cohere Labs Community Lead.
Currently working on mechanistic interpretability.
GitHub · LinkedIn ·
Cross-Lingual Emergent Misalignment
2026.05.03
Multilingual Induction Heads
2026.03.04
The Tokenization Tax
2026.02.24
A Non-pharmacological Recommendation System for Osteoporosis Management Using Large Language Models
2023.09.18